Skip to content

Generative AI engineering skills

LangGraph and LangChain for agent orchestration; Azure AI Foundry and OpenAI for models; Pinecone and Azure AI Search for retrieval; RAGAS and LangSmith for evaluation; and the data engineering stack underneath them.

Production tech stack

AI engineering tools I use in production.

How I choose the stack

The use case and constraints decide the tools. I prioritize grounded answers, traceable workflows, measurable evaluations, and an operating model the team can own.

Grounded retrievalAgent orchestrationLLM evalsTeam-ready deployment

GenAI & LLMs

01

LLM applications built with RAG, GraphRAG, agent workflows, and prompt design.

RAG and GraphRAGAgent workflowsPrompt design

Often includes

RAG ArchitectureGraphRAGKnowledge GraphsAgentic AIModel Context ProtocolLangGraphLangChainLlamaIndexHugging FaceOpenAIAzure OpenAILoRA/PEFTPrompt EngineeringBERT/T5AI AgentsA2A ProtocolHuman-in-the-Loop

Azure AI

02

Azure AI Foundry, Azure OpenAI, AI Search, Functions, and the data services around them.

Azure AI FoundrySearch and orchestrationTeam-ready deployment

Often includes

Azure AI FoundryCopilot StudioAzure AI SearchAzure FunctionsAzure Key VaultPromptFlowAzure KubernetesAzure Data LakeAzure DatabricksAzure Data Factory

Evaluation & Quality

03

Eval suites for retrieval relevance, groundedness, answer quality, and regression tracking.

LLM evaluationRetrieval qualityExperiment tracking

Often includes

BLEUROUGERecall@kNDCGF1 ScoreRAGASLangSmithLangfusePromptfoo

Vector Databases

04

Vector and graph-vector retrieval with embeddings, hybrid search, reranking, and semantic indexes.

Vector searchGraph-vector retrievalEmbedding strategy

Often includes

PineconeFAISSChromaDBPgvectorHybrid SearchGraph-Vector RetrievalSemantic SearchEmbeddings

LLMOps

05

Deployment, monitoring, orchestration, and automation for AI applications in production.

DeploymentMonitoringAutomation

Often includes

GitDockerKubernetesFastAPICI/CDAirflowRedisMLflowAzure DevOpsMonitoring

Programming & Data

06

Python, SQL, ML libraries, and data platforms used across model and data pipelines.

Python and MLData engineeringAnalytics workflows

Often includes

PythonPyTorchTensorFlowSQLPandasNumPyPySparkTypeScript/ReactKafkaData ModelingUnity CatalogDelta Lake

Keep exploring

AboutServicesProjectsExperienceCertificationsFAQBlogContact